Operations
Operations

Acknowledge risk

Mark a reliability **risk** as intentionally accepted. Use this only when a risk in the `<readiness>` block is genuinely a deliberate choice for this agent, not something to fix. The risk stops counting toward the "Could Improve" badge and moves to a collapsed "Acknowledged" list; it stays visible and auditable. **Errors can never be acknowledged**, they block publish for a reason; fix them instead. **Don't reach for this to silence a risk you could actually fix**, acknowledge means "correct as-is", not "ignore". Always confirm with the operator before acknowledging; a `reason` is mandatory and the operator should agree with it.

POST
/operations/acknowledge_risk

Authorization

bearerAuth
AuthorizationBearer <token>

A personal access token from Settings > API tokens.

In: header

Header Parameters

Idempotency-Key?string

A key you choose, such as a UUID, that makes this POST safe to retry. Within 24 hours the same key and request answers with the first successful answer and runs nothing, the same key with a different request answers 422 idempotency_key_reused, and a key whose request is still running answers 409 idempotency_key_in_use. A failed request keeps nothing, so the same key runs it again.

Length1 <= length <= 255

Request Body

application/json

TypeScript Definitions

Use the request body type in TypeScript.

Response Body

application/json

application/problem+json

application/problem+json

application/problem+json

application/problem+json

application/problem+json

application/problem+json

application/problem+json

curl -X POST "https://example.com/operations/acknowledge_risk" \  -H "Content-Type: application/json" \  -d '{    "riskKey": "string",    "reason": "string",    "agent": "string"  }'
{}
{
  "type": "http://example.com",
  "title": "string",
  "status": 0,
  "detail": "string",
  "code": "string",
  "errors": [
    {
      "path": "string",
      "message": "string"
    }
  ],
  "estimate": {},
  "retryAfterSeconds": 0
}
{
  "type": "http://example.com",
  "title": "string",
  "status": 0,
  "detail": "string",
  "code": "string",
  "errors": [
    {
      "path": "string",
      "message": "string"
    }
  ],
  "estimate": {},
  "retryAfterSeconds": 0
}
{
  "type": "http://example.com",
  "title": "string",
  "status": 0,
  "detail": "string",
  "code": "string",
  "errors": [
    {
      "path": "string",
      "message": "string"
    }
  ],
  "estimate": {},
  "retryAfterSeconds": 0
}
{
  "type": "http://example.com",
  "title": "string",
  "status": 0,
  "detail": "string",
  "code": "string",
  "errors": [
    {
      "path": "string",
      "message": "string"
    }
  ],
  "estimate": {},
  "retryAfterSeconds": 0
}
{
  "type": "http://example.com",
  "title": "string",
  "status": 0,
  "detail": "string",
  "code": "string",
  "errors": [
    {
      "path": "string",
      "message": "string"
    }
  ],
  "estimate": {},
  "retryAfterSeconds": 0
}
{
  "type": "http://example.com",
  "title": "string",
  "status": 0,
  "detail": "string",
  "code": "string",
  "errors": [
    {
      "path": "string",
      "message": "string"
    }
  ],
  "estimate": {},
  "retryAfterSeconds": 0
}
{
  "type": "http://example.com",
  "title": "string",
  "status": 0,
  "detail": "string",
  "code": "string",
  "errors": [
    {
      "path": "string",
      "message": "string"
    }
  ],
  "estimate": {},
  "retryAfterSeconds": 0
}

Accept agent prompt improvement POST

Accept an Optimise-Prompt proposal, publish its proposed system prompt as a new agent version. Use this after the optimise loop (optimise_agent_prompt) leaves a 'proposed' session you want to apply. It clones the current published version into a draft, writes the proposed prompt, and publishes it through the SAME path the editor uses (so the publish validator still gates it), then marks the session accepted. It only succeeds when the loop's re-run proved the proposed prompt raised the pass rate (pass_rate_after > pass_rate_before); a non-improving or already-resolved session is refused. This is the human-in-the-loop accept the self-improving loop deliberately gates on, never publish a prompt change blindly. Returns the new published version number. The session must belong to the current workspace and agent.

Add dashboard plot POST

Add one plot (widget) to an Overview dashboard from a natural-language description. Given a dashboardId (from list_dashboards) and a description like 'success rate by agent', this maps the request to a valid plot spec, data field, chart type, aggregation and group-by, all constrained to the dashboard's supported columns, and saves it to the dashboard's grid. To narrow the calls a plot counts (for example only calls where the caller hung up), pass `filters`; the description alone never filters. The dashboard's time range sets the dates; the description does not. Use it to build out a dashboard conversationally instead of the manual Add-Plot modal. This charts our OWN historical call data; it never changes how an agent behaves. Persists immediately; one plot per call. One short AI call drafts the plot and is billed to the workspace as AI usage.